Environ. Sci. Technol. 1991, 25, 777-782
(16) Leaderer, B. P. Air pollution from kerosene space heaters. Science 1982,218, 1113-1115. (17) Maxwell, Tobacco Int. January 1987. (18) Bluyssen, P.; Van De Loo, H.; Leaderer, B. P. Chamber and field studies of respirable suspended particulate deposition rates indoors. Proceedings of the 4th International Conference on Indoor Air Quality and Climate;Institute for Water, Soil and Air Hygiene: Berlin, 1987; Vol. 1, pp 549-553. (19) An investigation of infiltration and indoor air quality. Final report to the New York State Energy Research and Development Authority prepared by the Research Triangle Institute, January 1990. (20) Palmes, E. D.; Tomczyk, C.; March, A. W. Relationship of indoor NOz concentrations in use of unvented gas appliances. J. Air Pollut. Control Assoc. 1979, 29, 392. (21) Reiszner, K. D.; West, P. W. Collection and determination of sulfur dioxide incorporating permeation and West-Gaeke procedure. Environ. Sci. Technol. 1973, 7, 526. (22) Dietz, R. N.; Cote, E. A. Air infiltration measurements in a home using a convenient perfluorocarbon tracer technique. Enuiron. Int. 1982, 8, 419. (23) Turner, W. A.; Marple, V. A.; Spengler, J. D. Indoor aerosol impactor. Presented a t Third International Conference on Indoor Air Quality and Climate, Stockholm, Sweden, August 20-24, 1984. (24) Leaderer, B. P.; Koutrakis, P.; Briggs, S. L. K.; Rizzuto, J. The mass concentration and elemental composition of indoor aerosols in Suffolk and Onondaga Counties, New
York. Submitted for publication in Atmos. Environ. (25) Leaderer, B. P.; Cain, W. S.; Isseroff, R.; Berglund, L. G. Ventilation requirements in buildings. 11. Particulate matter and carbon monoxide from cigarette smoking. Atmos. Environ. 1984, 18, 99-106. (26) Leaderer, B. P.; Cain, W. S.; Isseroff, R.; Berglund, L. G. Tobacco smoke in occupied spaces: Ventilation requirements. Proc.-APCA, 74th Annu. Meet. 1981, 81-22.6, 1-14. (27) Schenker, M. B.; Samuels, S. J.; Kado, N. Y.; Hammond, S. K.; Smith, T. J.; Woskie, S. R. Markers of exposure to diesel exhaust in railroad workers. Research Report 33, Health Effects Institute: Cambridge, MA, 1990. (28) Miesner, E. A,; Rudnick, S. N.; Hu, F.; Spengler, J. D.; Preller, L.; Ozkaynak, H.; Nelson, W. Particle and nicotine sampling in public facilities and offices. JAPCA 1989, 39, 1577-1582. (29) Muramatsu, M.; Umemura, S.; Okada, T.; Tomia, H. Estimation of personal exposure to tobacco smoke with a newly developed nicotine personal monitor. Environ. Res. 1984, 35, 218. Received f o r review August 20, 1990. Accepted November 26, 1990. This work is supported by EPA cooperative Agreements CR-814150, CR-813594, and CR-813610. The field portion of this study was conducted by the Research Triangle Institute ( R T I )f o r the New York State Energy Research Development Authority (NYSERDA).
Oxidation of Chlorobenzene with Fenton's Reagent David L. Sedlak and Anders W. Andren" Water Chemistry Program, University of Wisconsin, Madison, Wisconsin 53706
w The degradation of chlorobenzene and its oxidation products by hydroxyl radicals generated with Fenton's reagent was studied. In the absence of oxygen, chlorophenols, dichlorobiphenyls (DCBs), and phenolic polymers were the predominant initial products. In the presence of oxygen, DCB yields decreased markedly and chlorobenzoquinone was also formed. Chlorophenol isomers were further oxidized by OH's to form chlorinated and nonchlorinated diols. DCBs and the phenolic polymers were also oxidized. The highest yield of product formed per mole of H 2 0 zconsumed was observed in the pH range of 2-3. The pH dependence and product distributions suggest that complexes of aromatic intermediate compounds with iron and oxygen may play a role in regulating reaction pathways. A t pH 3.0, approximately 5 mol of Hz02/mol of chlorobenzene were required to remove all of the aromatic intermediate compounds from solution. Introduction
Chlorinated aromatic hydrocarbons (CAHs) have been introduced to the environment from a variety of sources. Many of these compounds do not readily degrade and pose a threat to biota and human populations. Concern about the potential hazards associated with these compounds has resulted in laws and policies that require the cleanup of contaminated soil, sediments, surface water, and wastewater ( I ) . Treatment of these wastes necessitates the development of technologies to effectively degrade many types of CAHs. One potentially important method of destroying CAHs is through chemical oxidation by hydroxyl radicals generated with Fenton's reagent ( 2 , 3 ) . Fenton's reagent is 0013-936X/91/0925-0777$02.50/0
a mixture of hydrogen peroxide and ferrous iron (Fez+), which produces OH's according to reaction 1 ( 4 ) : Fez+ + H20, Fe3+ OH- + OH' (1)
-
The OH's produced in reaction 1 are capable of reacting with a variety CAHs (5-7). Preliminary studies for the design of waste treatment systems employing Fenton's reagent (2,3,8-10) indicate that the reaction is effective in the degradation of phenols, chlorophenols, formaldehyde, and octachloro-p-dioxin. However, none of these studies have focused on either the nature of intermediate products or factors affecting product yields and distributions. Furthermore, mechanistic studies on CAHs are mainly limited to systems in which oxygen is excluded and high H 2 0 2and substrate concentrations are present. Understanding the reaction mechanism for the oxidation of CAHs under conditions relevant to waste treatment is an essential step in the design of efficient, cost-effective Fenton's reagent treatment systems. These factors are especially important for this oxidation system because product yields and distributions may be drastically affected by environmental conditions such as pH and oxidant concentrations (11,12). Furthermore, identification and quantification of intermediate products is important because hydroxylated aromatic and dimeric intermediates may be recalcitrant and/or toxic. In this study we have evaluated reactions of Fenton's reagent with chlorobenzene and its intermediate oxidation products as a function of pH, and in the presence or absence of oxygen. Through determination of intermediate products and the effect of environmental variables on product yields and distributions, we have identified possible reaction mechanisms and optimal conditions for
0 1991 American Chemical Solciety
Environ. Sci. Technol., Vol. 25, No. 4, 1991
777
degrading these compounds. In addition, we have collected data that provide a basis for estimating the quantities of H 2 0 zrequired to oxidize all of the CAHs and aromatic intermediates.
Materials and Methods The following chemicals were obtained from Aldrich at the specified purities and were used without further purification: benzoquinone (97%), catechol (99+ 70), chlorobenzene (99+ 701, chlorohydroquinone (technical grade), 2-chlorophenol (99+ % ), 3-chlorophenol (98%), 4-chlorophenol (99+ % ),4-chlororesorcinol (98%), hydrogen peroxide (ACS reagent grade, 30% solution), ferric sulfate pentahydrate, ferrous sulfate heptahydrate (99+ %), 2,4,6-tri(2-pyridyl)-1,3,5-triazine (TPTZ) (97%). Dichlorobiphenyl isomers were purchased from Cambridge Isotope Laboratories. Aqueous solutions of various organic compounds, used for oxidation reactions and as HPLC standards, were prepared with Millipore Milli-Q water. Nitrogen was bubbled through the water for at least 1h prior to addition of solute for solutions used in reactions conducted in the absence of oxygen. Chlorobenzene solutions were stirred for a t least 1week prior to use. Solutions of chlorophenols and all other standards were prepared by dissolving the appropriate quantity in water and stirring a t least 1 h. Chlorobenzoquinone standards were prepared by oxidizing a chlorohydroquinone solution with excess Fe3+. Completion of the reaction was confirmed by the disappearance of the chlorohydroquinone peak on the HPLC chromatogram. Fenton's reagent reactions were conducted in a 500-mL round-bottom flask. The flask was maintained at 25 "C with a recirculating water bath and shielded from light with aluminum foil. For reactions conducted in the absence of oxygen, a nitrogen atmosphere was maintained over the solution at a slight positive pressure and samples were withdrawn through a sampling port with a syringe. Samples were collected after 1 h unless otherwise noted. The initial solution pH was adjusted with hydrochloric acid for reactions a t pH values up to 3.0. Reactions at pH values of 4.7 and 7.0 were buffered with acetic acid and phosphoric acid buffers, respectively. Hydrogen peroxide solutions, maintained a t 25 "C with a recirculating water bath, were added to the reactor containing the solute and Fez+either in a total volume of 10 mL a t the start of the reaction or at a rate of 5.0 mL/h with a metering pump. Ferrous iron was quantified by the method of Collins et al. (13), which involves spectrophotometric measurements of the Fe2+/TPTZ complex. Polar compounds were quantified by HPLC using a 25-cm reverse-phase column (ODs-2; Alltech) and a mobile phase of methanol/water with 0.5 mM acetate buffer. The following gradient was used: 10% methanol (hold 4 rnin); ramp to 50% methanol by 15 min (hold 6 rnin). Compounds were detected a t 280 nm and quantification was based on comparison with standards, except in the case of 4-chlorocatechol, which was estimated on the basis of the measured extinction coefficient of catechol. Samples withdrawn from the reactor (5 mL) were also extracted three times with 10 mL of petroleum ether and 0.2 mL of saturated sodium chloride and analyzed for chlorobenzene by GC/MS (Hewlett-Packard 5890/5970). These extracts were then concentrated to 1 mL and analyzed quantitatively for dichlorobiphenyls and qualitatively for other compounds. Dichlorobiphenyls were quantified by using 2,4',5-trichlorobiphenyl as an internal standard. Chromatographic conditions were as follows: 30-m RT,-5 778
Environ. Sci. Technol., Vol. 25, No. 4, 1991
2.0
I rn m A
E
v
Chlorobenzene Total Chlorophenols Chlorobenzoquinone
1.4
00
02
0 4
0 6
08
1 3
1 2
1 4
16
1 8
Time ( q r s ) Figure 1. Oxidation of chlorobenzene with Fenton's reagent. Initial conditions: Fez+, 5.0 mM; pH, 3.0
I . 4,4'-DCB
A
2
5
:2
3C
02
04
36
T-e
08
1 2
12
I
1 4
(hrs )
Figure 2. Formation of three dichlorobiphenyls isomers during oxidation of chlorobenzene with Fenton's reagent. 2,2'-DCB, 2,3'-DCB, and 3,3'-DCB are not shown because concentrations were significantly smaller than those observed for the other isomers. Conditions: Fez+, 5.0 mM; pH, 3.0
capillary column (Restek); oven programmed to start a t 50 "C (hold 4 min) then ramp to 280 "C a t 5 "C/min; detection with scan mode in the 5C-350 mass/charge range. UV/vis spectrophotometric measurements were performed on a Varian DMS 80 spectrophotometer with a slit width of 3 nm. UV/vis scans were conducted at a scan rate of 100 nm/min. Dissolved organic carbon (DOC) was analyzed on an automatic TOC analyzer (Model 700, 01 Corp.).
Results Results from the oxidation of chlorobenzene by addition of Hz02a t a rate of 5 mM/h a t an initial pH of 3.0 are shown in Figures 1and 2 . Chlorobenzene and all intermediate products disappeared within the first 2 h of the reaction. The three chlorophenol isomers, chlorobenzoquinone, and the six dichlorobiphenyl isomers were identified by GC/MS (Table I) and quantified by either GC/MS or HPLC. The mass spectra and retention times for these compounds were identical with those of the standards and the mass spectra closely matched those in the NBS database. A group of compounds with retention times similar to the dichlorobiphenyl isomers were also observed. These peaks were tentatively identified as hydroxymonochlorobiphenyls or hydroxydichlorobiphenyls on the basis of the results of a computerized search of the NBS mass spectral library and similarity of their mass
Table 11. Chlorophenol a n d Dichlorobiphenyl Produced during Oxidation of Chlorobenzene with Fenton's Reagent with a n d without oxygen
Table I. Intermediate Compounds Identified from Oxidation of Chlorobenzene with Fenton's Reagent peak 1
2 3 4 5 6 7 8 9
10 11 12
compound
RTR
IDb
quant'
2-chlorophenol chlorobenzoquinone 3-chlorophenol 4-chlorophenol 2,2'-DCB 2,3'-DCB 2,4'-DCB 3,3'-DCB 3,4'-DCB 4,4'-DCB MHClBps DHClBps
11.9 10.7 15.9 17.0 21.2 25.3 25.8 28.7 28.8 29.4 manye manye
std std stdd stdd std std std std std std NBS NBS
HPLC HPLC HPLC HPLC GC/MS GC/MS GC/MS GC/MS GC/MS GC/MS NQf NQf
atmosn
orthob
metab
N2
0.073 (53.9) 0.041 (36.1)
0.026 (19.2) 0.023 (20.3)
O2
Retention time in minutes on column used for quantification. Method of GC/MS identification: std, matched with mass spectra and retention times of standard; NBS, compared with NBS mass spectral database and spectra described by Tulp et al. (14). Method used for quantification. These chlorophenol isomers coeluted on GC/MS. Mass spectra showed characteristics of both compounds. e Four hydroxymonochlorobiphenyl compounds (RT = 18.5, 22.5, 24.3, and 25.1 min) and four hydroxydichlorobiphenyl compounds (RT = 26.2, 26.9, 27.5, and 27.7 min) were identified. At least 10 additional mono- and dichlorinated hydroxylated biphenyls were also identified between 29 and 31 min but could not be fullv seearated. fNQ. not auantified.
concnSemM parab DCBs' 0.028 (20.7) 0.049 (43.2)
totald
0.0085 (6.3) 0.0005
0.1355 0.1135
(0.4)
"Initial conditions: Fez+,4.5 mM; pH, 3.0; H202, 2.7 mM; chlorobenzene, 3.8 mM. Results are average values for two experiments. Concentrations of chlorophenol isomers. 'Total concentration of dichlorobiphenyl (DCB) isomers. Total amount of chlorophenols and DCBs produced. e Values in parentheses represent percentage of total.
140
7
A
9 -
DOC
4
0.000
4
1
0
1
2
3
4
5
6
7
8
PH Flgure 4. Relationship of total chlorphenol yield and pH. Initial conditions: Fe", 5.0 mM; H202,2.7 mM; chlorobenzene, 3.8 mM. Error bars represent 95% confidence intervals based on pooled variances observed in duplicate experiments at each pH.
Ij0 8 L
.
tions of dichlorobiphenyls (DCBs) when compared to those in air. Furthermore, chlorophenol isomer distributions under nitrogen contained a higher proportion of 0chlorophenol than reactions performed with oxygen. When further experiments were conducted in the presence of oxygen (Figure 4), the highest yield of product formed per mole of H,O, consumed was observed a t pH values in the range of 2-3. Chlorophenol isomer distributions remained constant (combining the results of duplicate analyses a t each pH the following mean isomer distributions and 95% confidence intervals were calculated: ortho, 35.3 f 2.1%; meta, 19.7 f 2.6%; para, 45.0 f 1.7%) and DCB concentrations remained below lo4 M over the entire pH range. The oxidation of pure solutions of chlorophenol isomers by Fenton's reagent was also qualitatively evaluated (Figure 5). Chlorinated diols resulting from ortho or para substitution accounted for most of the observed products identified by GC/MS. Small quantities of nonchlorinated diols (hydroquinone, catechol, and benzoquinone) were also observed. Reactions with 2-chlorophenol and 3-chlorophenol exhibited the same broad-band absorption observed for chlorobenzene. Reactions with 4-chlorophenol passed through a green-colored stage prior to onset of broad-band absorption. Oxidation of a 4-chlorophenol solutions was followed quantitatively in the presence and absence of oxygen as depicted in Figure 6. In both cases, 4-chlorocatechol was observed as the predominant product. Compound identification was based upon analysis of mass spectra and comparison with HPLC chromatograms from Sehili et al. (16). Quantification was based upon extinction coefficient measurements for catechol and assumes that the UV Environ. Sci. Technol., Vol. 25, No. 4, 1991
779
isomer
products observed para
ortho'
& &
rneta
None detected
0
3-chlorocatechol
c h Io robe nzo q u inone
catechol
I
I None detected
0
3-chlorocatechol
4-chlorocatechol
I
c hlorobenzoquinone
OH
5'
0
0
4-chlorophenol
OH
OH
4-chlorocatechol
hydroquinone
0 0
benzoquinone
I
C1
@OH
OH 4-chlororesorcinoi
'Location of hydroxyl addition (relative to original OH group on chlorophenol isomer) Figure 5. Products observed from chlorophenol oxidation by Fenton's reagent. Initial conditions: Fez+, 0.5 mM; pH, 3.0; HO ,, isomer, 3.0 mM.
4.5 mM; chlorophenol
30
A
2
E
25
s v
Clcatechol; A i r 4-Clp; Nitrogen Clcatechol; Nitrogen
2.0
C
.-0
P
+ -
1.5
+ C 10
C
0 0.5
0.0 0.0
05
'.0
15
2.0
2.5
30
3.5
4.0
4.5
Time (hrs.) Figure 6. Oxidation of 4-chlorophenol with Fenton's reagent. Initial conditions: Fez+, 0.5 mM; H,O, addition rate, 3.0 mM/h; pH, 3.0. 4-Chiororesorcinol, hydroquinone, and benzoquinone were detected at concentrations less than 0.2 mM.
spectra of the chlorinated compound is similar to its nonhalogenated parent. Other intermediate compounds were not detected in significant quantities by GC/MS or HPLC. Results from these experiments also illustrate that oxygen almost doubles the rate of disappearance of 4chlorophenol.
Discussion The overall results of our experiments indicate that Fenton's reagent can effectively degrade chlorobenzene, chlorophenols, and dichlorobiphenyls. In the presence of oxygen at pH 3.0, approximately 5 mol of H202/molof chlorobenzene is required to completely remove the aromatic intermediates from solution. Results from DOC analysis indicate that the aromatic intermediates undergo ring cleavage prior to mineralization. Cessation of the reaction after approximately 4 h, as evidenced by stabilization of DOC concentrations, pH, and Fe2+, is most likely attributable the inability of many ring-cleavage intermediates to regenerate Fez+. Chlorobenzene Reactions. By analogy to previous work on OH' reactions with aromatic compounds and evaluation of our results, we believe that the oxidation of chlorobenzene by Fenton's reagent (Figure 7 ) follows a 780
Environ. Sci. Technol., Vol. 25, No. 4, 1991
Figure 7. Proposed reaction pathway for oxidation of Chlorobenzene with Fenton's reagent.
complicated reaction pathway in which products are formed via several different mechanisms controlled by factors such as oxidant concentrations and pH. The first step in the reaction sequence, OH' attack on chlorobenzene (reaction a), likely results in the formation of chlorohydroxycyclohexadienyl (ClHCD) radical I (7,17), which may undergo one of several possible further reactions. In the absence of oxygen, or other strong oxidants, the two predominant reactions are dimerization to produce dichlorobiphenyls (reaction 3) and bimolecular disproportionation to produce chlorophenol and chlorobenzene (reaction 4). Both of these reactions exhibit an overall stoichiometry of 2 mol of Hz02/molof chlorobenzene oxidized, which is consistent with our observations (see Table 11). A similar stoichiometry has been observed in the oxidation of benzene by Fenton's reagent, but biphenyl yields were as high as 50% (18). In the presence of oxygen, or other strong oxidants, different product distributions are observed because several additional reactions contribute to product formation. Reactions of the oxidant (0,) with the ClHCD radical (reactions 5 and 6) predominate because they are first
order with respect to ClHCD radical while bimolecular reactions 3 and 4 are second order with respect to the radical. Reaction of the ClHCD radical with O2results in lower DCB yields, different chlorophenol isomer distributions and higher chlorobenzoquinone yields than reactions under nitrogen. Reactions of ClHCD radicals with O2 are also important because they could result in the production of hydroperoxy radicals (via reaction 5 ) or H202 (19), which could further oxidize CAHs. The formation of chlorobenzoquinone as a result of oxygen attack on the ClHCD radical in a position para to the initial OH‘ addition was expected, based upon previous experimental results with benzene (19,20). However, the absence of ortho-substituted diols (catechol or chlorocatechol) observed in these reactions is unexpected given the ortho/para directing tendency of hydroxyl groups. One possible explanation is steric hindrance a t the ortho position caused by an iron-oxygen complex similar to that observed in the C u / 0 2 analogue of Fenton’s reagent (20). The formation of chlorobenzoquinone by a pathway involving O2 rather than via a secondary reaction of a chlorophenol isomer is strongly supported by our results. The decreased relative yields of o-chlorophenol observed in the presence of oxygen can be explained by a more favorable reaction of O2 with ortho ClHCD radicals than other ClHCD radicals. Furthermore, catechol and chlorocatechol isomers were produced as primary products in reactions with solutions of the chlorophenol isomers and would presumably be present if chlorobenzoquinone was a secondary reaction product. Chlorophenol Reactions. Hydroxyl radical attack on the chlorophenol isomers is directed by the position of the hydroxyl group, which is a stronger ortho/para director than chlorine (22) (Figure 5 ) . Reactions performed with each of the chlorophenol isomers resulted almost exclusively in products of OH’ attack at a position ortho or para to the hydroxyl group. Only one product of OH’ attack a t a position meta to the hydroxyl group (4-chlororesorcinol) was observed. The formation of nonchlorinated products indicates that the presence of a chlorine group did not prevent OH’ attack on the ring. Hydroxyl radical attack did however occur more readily a t positions that were not occupied by chlorine groups because dechlorination occurs via a different reaction mechanism, which is probably not as efficient as reactions not resulting in dechlorination. The visible light absorption and our inability to account for all of the reaction products [also observed by Kunai et al. (1911 suggest that some compounds are formed that are not amenable to HPLC analysis. The green color observed during the oxidation of 4-chlorophenol is most likely attributable to the formation of a complex of Fe3+ with one of the aromatic oxidation products. This green color was also observed when ferric sulfate and catechol were mixed a t concentrations comparable to those observed in our experiments. Furthermore, a variety of other colored complexes have been reported when other disubstituted aromatic compounds were mixed with Fe3+ (23). Polymerization Reactions. The broad-band visible light absorption exhibited during the oxidation of chlorobenzene and the chlorophenol isomers, which has been observed previously with benzene and phenol (8, 19),is most likely attributable to the formation of phenolic polymers. Formation of phenolic polymers in reactions involving OH’ was noted by Stein and Weiss (24) and has been implicated in the formation of humic materials (25-27). Hydroxychlorobiphenyls have been identified previously during the radical polymerization of 4-chloro-
phenol initiated by reactions with free chlorine radicals (25). The compounds we observed by GC/MS may result from the polymerization of phenoxy radicals or ClHCD radical cations. Alternatively, these compounds may have been produced via a secondary reaction of OH’ with the dichlorobiphenyls, as observed in the oxidation of chlorinated biphenyls by Fenton’s reagent (15). Disappearance of the broad-band absorption and the dimers as the reaction progressed suggests that these polymers and dimers are ultimately amenable to oxidation. pH Dependence. Optimal product yields (in terms of the amount of chlorobenzene oxidized per mole of H 2 0 2 consumed) in the pH range of 2-3 have been reported previously for phenol (8). One possible explanation of the pH effect is acid-catalyzed base-catalyzed elimination of water from ClHCD radicals followed by reduction to chlorobenzene (reactions 7 and 8) (18). However, this reaction only contributes to a 50% decrease in product yield for oxidation of benzene in the absence of oxygen at pH 1 (18). In the presence of oxygen, the reaction should exert even less of an effect because the unimolecular elimination of water is competing with the rapid oxidation of ClHCD by 02. Another possible explanation for the pH effect is participation of substrate-Fe-02 or substrate-Fe-H202 complexes in the reaction. Kinetic results from oxidation of diols in the Fenton’s reagent system (28, 29) and in the presence of Mn2+/02(30) do not follow predictions for a radical mechanism, but are better explained by the formation of organometallic ternary complexes. Futhermore, the stability constants for such complexes would likely favor their formation a t the pH values where optimal product yields were observed, because ferric complexes with aromatic ligands are most readily formed in the pH range of 2-4 (31). Therefore, higher product yields in this pH range could be explained by a reaction involving the organometallic complex where either H 2 0 2is regenerated (for example, proton abstraction from chlorohydroquinone by 0,) or through increased rates of reactions in which H 2 0 2is wasted.
Conclusions Fenton’s reagent can be employed to effectively degrade recalcitrant CAHs such a chlorobenzene and chlorophenol. The most direct mechanism for CAH degradation proceeds through hydroxylation followed by ring cleavage and mineralization. Another possible reaction pathway involves the formation of dimers (such as chlorinated biphenyls and hydroxychlorobiphenyls)and colored aromatic polymers, which are oxidized by subsequent OH’ attack. Complete mineralization of organic compounds was not observed because some of the ring-cleavage products were unable to reduce Fe3+. If necessary, these ring-cleavage products could be futher oxidized by addition of more Fe2+, as demonstrated in the oxidation of phenol with the electrolytic Fenton’s reagent system (8). Evaluation of factors affecting the reaction mechanism has also helped to define reaction conditions resulting in the greatest loss of CAHs per mole of H202consumed. The presence of oxygen or other strong oxidants favors the more direct oxidation pathway and also follows a stoichiometry in which less H 2 0 2is required to degrade the CAHs. The reaction also follows a pH dependence and is most efficient in the pH range of 2-3. Registry No. 2,2’-DCB, 13029-08-8; 2,3’-DCB, 25569-80-6; 2,4’-DCB, 34883-43-7; 3,3’-DCB, 2050-67-1; 3,4’-DCB, 2974-90-5; 4,4’-DCB, 2050-68-2; chlorobenzene, 108-90-7; 2-chlorophenol, 95-57-8; chlorobenzoquinone, 695-99-8; hydroxymonochlorobiphenyl, 132178-75-7;hydroxydichlorobiphenyl,53813-74-4; oxygen, Environ. Sci. Technol., Vol. 25, No. 4, 1991
781
Environ. Sci. Technol. 1991, 25,782-788
7782-44-7; 3-chlorophenol, 108-43-0;4-chlorophenol, 106-48-9.
Literature Cited Porter, J. W. Chem. Eng. Prog. 1989, 85(4), 16-25. Barbeni, M.; Pelizzetti, E.; Borgarello, E.; Serpone, N. Chemosphere 1987, 16, 2225-2237. Bowers, A. R.; Eckenfelder, W. W.; Gaddipati, P.; Monsen, R. M. Water Sei. Technol. 1989, 21, 477-486. Merz, J. H.; Waters, W. A. J. Chem. SOC.1949, 2427-2433. Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J . Phys. Chem. Ref. Data 1988, 17, 513-886. Farhataziz; Ross, A. B. Selected Specific Rates of Reactions o f Transients from Water in Aqueous Solution. III. Hydroxyl Radical and Perhydroxyl Radical and Their Radical Zons; Report NSRDS-NBS 59; National Bureau
of Standards: Washington, DC, 1977. Dorfman, L. M.; Taub, I. A.; Buhler, R. E. J. Chem. Phys. 1962, 36, 3051-3061. Sudoh, M.; Kodera, T.; Sakai, K.; Zhang, J. Q.;Koide, K. J . Chem. Eng. Jpn. 1986, 6 , 513-518. Murphy, A. P.; Boegli, W. J.; Price, M. K.; Moody, C. D. Environ. Sei. Technol. 1989, 23, 166-169. Watts, R. J.; Miller, G. C.; Smith, B. R.; Rauch, P. A,; Tyre, B. W. Division of Environmental Chemistry, American Chemical Society, Miami Beach, Florida, September 1C-15, 1989; Extended Abstract, pp. 346-349. Walling, C. Acc. Chem. Res. 1975, 8, 125-131. Jefcoate, C. R. E.; Lindsay Smith, J. R.; Norman, R. 0. C. J . Chem. Soc. B 1969, 1013-1018. Collins, P. F.; Diehl, H.; Smith, G. F. Anal. Chem. 1959, 31, 1862-1867. Tulp, M. T. M.; Olie, K.; Hutzinger, 0. Biomed. Mass Spectrom. 1977, 4 , 310-316. Sedlak, D. L.; Andren, A. W. In preparation. Sehili, T.; Bonhomme, G.; Lemaire, J. Chemosphere 1988, 17, 2207-2218.
(17) Eberhardt, M. K.; Yoshida, M. J . Phys. Chem. 1973, 77, 589-597. (18) Walling, C.; Johnson, R. A. J . Am. Chem. Soc. 1975, 97, 363-367. (19) Kunai, A.; Hata, S.; Ito, S.; Sasaki, K. J . Am. Chem. Soc. 1986, 108, 6012-6016. (20) Ito, S.; Kunai, A.; Okada, H.; Sasaki, K. J . Org. Chem. 1988, 53, 296-300. (21) Groves, J. T.; Van Der Puy, M. J . Am. Chem. SOC.1974, 96, 5274-5275. (22) Fessenden, R. J.; Fessenden, J. S. Organic Chemistry; PWS Publishing: Boston, MA, 1982. (23) Schofield,P. J.; Ralph, B. J.; Green, J. H. J . Phys. Chem. 1964, 68, 472-476. (24) Stein G.; Weiss, J. J . Chem. Soc. 1951, 3265-3274. (25) Voudrias, E. A.; Larson, R. A,; Snoeyink, V. L. Enuiron. Sci. Technol. 1985, 19, 441-449. (26) Shindo, H.; Huang, P. M. Nature 1982, 298, 363-366. (27) Larson, R. A.; Hufnal, J. M. Limnol. Oceanogr. 1980,25, 505-5 12. (28) Eckschlager, K.; Veprek-Siska, J. Collect. Czech. Chem. Commun. 1972, 37, 1623-1634. (29) Eckschlager, K.; Horsak, I.; Veprek-Siska,J. Collect. Czech. Chem. Commun. 1974, 39, 2353-2362. (30) Tyson, C. A.; Martell, A. E. J . Am. Chem. Soc. 1972, 94, 939-945. (31) Martell, A. E.; Smith, R. M. Critical Stability Constants; Plenum: New York, 1977; Vol. 3. Received for review March 12,1990. Revised manuscript received October 8, 1990. Accepted December 4, 1990. This work was funded by the U S . Air Force Officeof Scientific Research, Grant AFOSR-88-0301. Funding was also obtained from the University of Wisconsin Sea Grant College Program under grants from the Office of Sea Grant, NOAA, U S . Department of Commerce,and the State of Wisconsin (Federal Grant NA84AA-D-00065).
Physical Factors Influencing Winter Precipitation Chemistry Jeffrey L. Collett, Jr.,+ Andri S. H. Privet, Johannes Staehelin," and Albert Waldvogel Atmospheric Physics ETH, Honggerberg HPP, 8093 Zurich, Switzerland
Two case studies of winter precipitation events highlight the roles of transport and snow crystal riming (the capture of supercooled cloud droplets by snow crystals) in determining precipitation chemistry. In one case, passage of a cold front leads to a change in the air mass producing precipitation over the monitoring site. A simultaneous decrease in precipitation ion concentrations is observed. Correlations of the ion concentrations with the pseudoequivalent potential temperature, which serves as an air mass identifier, suggest that the decrease in ion concentrations is caused by the air mass change, rather than by washout of aerosols and gases from the atmosphere. In the second case, evidence is presented indicating that an increase in precipitation ion concentrations results from significant capture of polluted cloudwater droplets by the snow crystals. Influences on precipitation chemistry from both processes, transport and riming, can be large. In order to study other processes influencing precipitation chemistry that occur on similar time scales (minutes to hours), such as aerosol and gas scavenging or aqueousphase oxidation, it is important to evaluate possible confounding effects of tranmort and riming.
Introduction The chemical composition of precipitation is determined 'Present address: Institute for Environmental Studies, 1101 W. Peabody Dr., University of Illinois at Urbana-Champaign, Urbana, IL 61801. 782
Environ. Sci. Technol., Vol. 25, No. 4, 1991
through the interaction of numerous chemical and physical processes. Some of the more commonly studied of these include nucleation and impaction scavenging of aerosol particles and scavenging of soluble gases within the cloud (1-41, below-cloud scavenging of aerosol particles and soluble gases by precipitation particles (4-a), and chemical reactions within the aqueous phase (9, IO). Our experience in studying aerosol scavenging in winter precipitation systems (4, 7) has revealed great variety between individual events. Processes primarily responsible for controlling precipitation chemistry in one case may play a subordinate role in another event. In this paper we will focus on two frequently ignored physical processes that we have observed playing an important role in determining the chemistry of winter precipitation in central Switzerland: an air mass change and the capture of supercooled cloud droplets by snow crystals (riming). Two case studies will be used to illustrate the influence these processes exert on precipitation chemistry. One of the weaknesses of ground-based precipitation studies is their limited ability to provide information about the temporal evolution of parameters of interest in a moving air mass. Because the air being sampled at a fixed station is changing with time, evaluating processes like aqueous-phase sulfur oxidation or aerosol scavenging can be greatly complicated by a change in background conditions that occurs on a similar time scale. This situation may be particularly problematic during frontal passage (4,
0013-936X/91/0925-0782$02.50/0
0 1991 American Chemical Society
